http://www.npr.org/sections/ed/2016/04/16/473273571/why-teachers-need-to-know-the-wrong-answers
For example, Sadler and colleagues created a high school astronomy course. In one of the lessons, students looked at pictures of the sun taken through the same telescope at each month of the year. Most predicted that the sun would appear larger in the hot months. However, once they got out the rulers, they would discover that the sun is biggest (i.e., closest) in January. (The closest point in our orbit, the "perihelion", was January 2 this year.)
This example caught my attention. Sure, the kids' assumption was incorrect, but if they were in the southern hemisphere their assumption would be correct but only a coincidence, not the reason.